Skip to content

Nous Research Launches Hermes Business for Team-Based AI Agent Workflows

The new Nous Portal tier gives organizations shared balances, member-level spending controls, and team-wide Hermes skills, while Enterprise adds private deployment options.

Table of Contents

Nous Research has launched Hermes Business, a new team account option for Hermes Agent that lets organizations invite colleagues, share one central balance, set per-member usage caps, and build reusable AI skills across a shared workspace.

The release expands Hermes beyond individual agent use into a more structured product for companies that want central billing, access controls, shared automation, and deployment options. Teams can use agents across multiple channels while turning repeated workflows into shared organizational capability.

Nous is also offering Hermes Enterprise for organizations that want the same agent system deployed on-premises or in a private cloud environment.

Hermes Business Brings Team Accounts to Nous Portal

Hermes Agent is Nous Research’s open-source autonomous agent, built to run across local machines, cloud environments, servers, and messaging channels. The agent is designed around persistent memory, reusable skills, scheduled automations, browser and web tooling, model access, and the ability to operate through interfaces such as CLI, Telegram, Discord, Slack, WhatsApp, Signal, email, and other gateways.

Hermes Business adds an organizational layer to that system through Nous Portal, the account and billing hub that powers Hermes model access, tools, and cloud hosting. Instead of each user running Hermes as a separate individual setup, a team can create a shared business account and invite members into the same environment.

Within that shared account, teams can use a shared credit balance while assigning per-member caps, giving companies more control over how agent usage is funded and distributed, giving teams a way to experiment with agents without giving every employee a disconnected billing setup.

The product is centered around four team-oriented capabilities: trace ownership, shared skills, spending intelligence, and proprietary workflow development. Hermes lets organizations own their agentic work across models, automatically share useful skills discovered by team members, monitor AI cost and return on investment, and build internal memory, skills, and workflows that accumulate into company-specific capability.

Shared Skills Are the Main Product Bet

Hermes will improve for teams through repeated organizational use.

Hermes already includes a skills system, which allows the agent to create and reuse procedural knowledge from prior tasks. In this new business context, that means a workflow discovered by one person can become available to others in the same organization. A sales operations team, for example, could standardize a research routine; an engineering team could preserve a deployment checklist; a content team could reuse a production workflow across multiple projects.

That shared-skill model is how, as Nous put it, skills can “compound into proprietary IP.” The value comes from the team’s preferred operating patterns becoming reusable agent instructions over time; the agent becomes more useful as it records how a team works, which tools it uses, and which workflows produce repeatable, optimal results.

Enterprise Adds Private Infrastructure Options

Nous is also introducing Hermes Enterprise for customers that need more control over where the agent runs. Hermes Enterprise can run on infrastructure controlled by the customer, with tailored deployment, single sign-on, service-level agreements, and onboarding support.

The product is intended to provide a “sovereign AI stack,” affording deployment options (available for for on-premises or private cloud environments) that matter most to larger companies where agent access to internal systems, documents, code, or customer data can raise security, compliance, and procurement requirements that are difficult to satisfy with a standard cloud-only product.

Comments

Latest